# Codebook

## Tweet-Level Data

* `id`: Tweet IDs.
* `date`: Date when a tweet was posted.
* `pred`: Indicating whether a tweet is about vaccine mandates (0 or 1).
* `vader_neg`: Negativity level from `VADER` (range: 0 to 1).
* `vader_neu`: Neutrality level from `VADER` (range: 0 to 1).
* `vader_pos`: Positivity level from `VADER` (range: 0 to 1).
* `vader_compound`: Compound sentiment score from `VADER` (range: -1 to 1).
* `NRCLex_anger`: Proportion of words categorized as anger in the NRC Lexicon (range: 0 to 1).
* `NRCLex_anticipation`: Proportion of words categorized as anticipation in the NRC Lexicon (range: 0 to 1).
* `NRCLex_disgust`: Proportion of words categorized as disgust in the NRC Lexicon (range: 0 to 1).
* `NRCLex_fear`: Proportion of words categorized as fear in the NRC Lexicon (range: 0 to 1).
* `NRCLex_joy`: Proportion of words categorized as joy in the NRC Lexicon (range: 0 to 1).
* `NRCLex_sadness`: Proportion of words categorized as sadness in the NRC Lexicon (range: 0 to 1).
* `NRCLex_surprise`: Proportion of words categorized as surprise in the NRC Lexicon (range: 0 to 1).
* `NRCLex_trust`: Proportion of words categorized as trust in the NRC Lexicon (range: 0 to 1).
* `nlp_sentiment`: Sentiment classification of a tweet as positive, negative, or neutral by TweetNLP.
* `nlp_pos_label`: Indicating  whether a tweet is classified as positive by TweetNLP (0 or 1).
* `nlp_neu_label`: Indicating  whether a tweet is classified as neutral by TweetNLP (0 or 1).
* `nlp_neg_label`: Indicating  whether a tweet is classified as negative by TweetNLP (0 or 1).
* `nlp_anger`: Indicating  whether a tweet expresses anger as classified by TweetNLP (0 or 1).
* `nlp_anticipation`: Indicating  whether a tweet expresses anticipation as classified by TweetNLP (0 or 1).
* `nlp_disgust`: Indicating  whether a tweet expresses disgust as classified by TweetNLP (0 or 1).
* `nlp_fear`: Indicating whether a tweet expresses fear as classified by TweetNLP (0 or 1).


## State-Level Panel Data

* `state`: U.S. state.
* `date`: Date of observation.
* `p_man`: Proportion of mandate-related tweets.
* `vader_neg`: Average negativity score of vaccine-related tweets according to `VADER`.
* `NRC_anger`: Average proportion of anger-related words in vaccine-related tweets based on the NRC Lexicon.
* `nlp_neg_label`: Proportion of vaccine-related tweets classified as negative by TweetNLP.
* `nlp_anger`: Proportion of vaccine-related tweets classified as expressing anger by TweetNLP.
* `word_freedom`: Proportion of vaccine-related tweets containing freedom-related words.
* `vader_neg.cdc`: Average negativity score of tweets about public health officials according to `VADER`.
* `NRC_anger.cdc`: Average proportion of anger-related words in tweets about public health officials based on the NRC Lexicon.
* `nlp_neg_label.cdc`: Proportion of tweets about public health officials classified as negative by TweetNLP.
* `nlp_anger.cdc`: Proportion of tweets about public health officials classified as expressing anger by TweetNLP.
* `log_new_case_7ma`: Logarithm of the seven-day moving average of new COVID-19 cases.
* `log_new_death_7ma`: Logarithm of the seven-day moving average of new COVID-19 deaths.
* `log_new_dose_7ma`: Logarithm of the seven-day moving average of new COVID-19 vaccine doses.
* `Series_Complete_Pop_Pct`: Percentage of the population that has completed the COVID-19 vaccine series.
* `Series_Complete_Pop_Pct_sd`: Standard deviation of the percentage of the population that has completed the COVID-19 vaccine series across counties within a state.

